SketchLog

Status Python Version License

Bounded-memory streaming metrics with explicit approximation guarantees.

SketchLog allows you to ingest high-throughput event streams and extract accurate percentiles and cardinalities in constant memory. It combines DDSketch for latencies, HyperLogLog for unique items, and Count-Min Sketch for event frequencies.

Instead of storing arrays of events or exporting raw telemetry, SketchLog compresses the statistical shape of your data in real time, making it ideal for continuous monitoring, edge devices, and memory-constrained environments.

📚 Documentation

The full documentation is available at SketchLog Documentation Site (or via /docs in the repository). It includes:

Architecture: Details on distributed merges, drift detection, and memory footprint.
Benchmarks: Memory footprint scaling and CPU throughput limits.
Guarantees: Published error bounds, implementation preconditions, and checked capacity limits.
Integrations: How to integrate with Prometheus, FastAPI, and OpenTelemetry.
API Reference: Core classes, methods, and persistence operations.
Contributing: Guidelines for contributing to SketchLog.

Local Documentation Build

To build and preview the documentation locally:

pip install ".[docs]"
mkdocs serve

Installation

1. The Core Library (Python)

If you just want to use the high-performance sketching data structures directly in Python:

pip install sketchlog

2. The Standalone Server

To run SketchLog as an independent network service (like Redis or Prometheus):

pip install "sketchlog[server]"
uvicorn sketchlog.server:app --port 8080

Quickstart

🐍 Python (Embedded Library)

from sketchlog import StreamLog

log = StreamLog()
log.add_latency(42.5)
log.add_batch([15.0, 88.2, 42.1])
log.add_unique("user_12345")
log.add_event("cache_miss", count=5)

print(f"p99 Latency:  {log.p99():.2f}ms")

🟨 TypeScript / Node.js SDK

Connect to the standalone server via HTTP:

npm install @sketchlog/client

import { SketchLogClient } from '@sketchlog/client';

const client = new SketchLogClient({ endpoint: 'http://localhost:8080' });

// Validated HTTP ingestion with bounded retries
await client.ingestEvents('production_api', {
  latencies: [42.5, 15.0, 88.2, 42.1],
  uniques: ["user_12345"],
  events: { "cache_miss": 5 }
});

🐹 Go SDK

go get github.com/SBALAVIGNESH123/sketchlog/clients/go@v1.2.3

import "github.com/SBALAVIGNESH123/sketchlog/clients/go"

client := sketchlog.NewClient(sketchlog.ClientOptions{
    Endpoint: "http://localhost:8080",
})

batch := sketchlog.EventBatch{
    Latencies: []float64{42.5, 15.0, 88.2, 42.1},
    Uniques:   []string{"user_12345"},
    Events:    map[string]int64{"cache_miss": 5},
}

err := client.IngestEvents(ctx, "production_api", batch)

Distributed Clustering (Beta)

SketchLog supports multi-node clustering without requiring external coordination services like Redis.

You can run a cluster using the following environment variables:

SKETCHLOG_NODE_ID="node-1"
SKETCHLOG_PEERS="http://node2:8000,http://node3:8000"
SKETCHLOG_CLUSTER_SECRET="your-secret-token"
uvicorn sketchlog.server:app --port 8000

Performance trade-off: Mesh mode currently uses deterministic Python sketches to guarantee identical merge state across heterogeneous nodes. Measure this mode with your own traffic before capacity planning.

Community

Join us in Slack! Join SketchLog Slack

What this is not

SketchLog is a streaming metrics compression layer. It is deliberately not:

Not a tracing system. No request paths, no correlation IDs, no causal chains.
Not a time-series database. No historical drill-down, no label indexing.
Not an observability platform. No raw log storage, no ad-hoc queries.
Not exact. All results are probabilistic with bounded error. If you need exact percentiles, use numpy.

MIT License

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

sketchlog-1.2.3.tar.gz (136.9 kB view details)

Uploaded Jun 30, 2026 Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

sketchlog-1.2.3-cp314-cp314-win_amd64.whl (410.7 kB view details)

Uploaded Jun 30, 2026 CPython 3.14Windows x86-64

sketchlog-1.2.3-cp314-cp314-musllinux_1_2_x86_64.whl (1.3 MB view details)

Uploaded Jun 30, 2026 CPython 3.14musllinux: musl 1.2+ x86-64

sketchlog-1.2.3-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (247.7 kB view details)

Uploaded Jun 30, 2026 CPython 3.14manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

sketchlog-1.2.3-cp314-cp314-macosx_11_0_arm64.whl (221.7 kB view details)

Uploaded Jun 30, 2026 CPython 3.14macOS 11.0+ ARM64

sketchlog-1.2.3-cp314-cp314-macosx_10_15_x86_64.whl (229.3 kB view details)

Uploaded Jun 30, 2026 CPython 3.14macOS 10.15+ x86-64

sketchlog-1.2.3-cp313-cp313-win_amd64.whl (400.4 kB view details)

Uploaded Jun 30, 2026 CPython 3.13Windows x86-64

sketchlog-1.2.3-cp313-cp313-musllinux_1_2_x86_64.whl (1.3 MB view details)

Uploaded Jun 30, 2026 CPython 3.13musllinux: musl 1.2+ x86-64

sketchlog-1.2.3-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (247.6 kB view details)

Uploaded Jun 30, 2026 CPython 3.13manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

sketchlog-1.2.3-cp313-cp313-macosx_11_0_arm64.whl (221.6 kB view details)

Uploaded Jun 30, 2026 CPython 3.13macOS 11.0+ ARM64

sketchlog-1.2.3-cp313-cp313-macosx_10_13_x86_64.whl (229.5 kB view details)

Uploaded Jun 30, 2026 CPython 3.13macOS 10.13+ x86-64

sketchlog-1.2.3-cp312-cp312-win_amd64.whl (400.4 kB view details)

Uploaded Jun 30, 2026 CPython 3.12Windows x86-64

sketchlog-1.2.3-cp312-cp312-musllinux_1_2_x86_64.whl (1.3 MB view details)

Uploaded Jun 30, 2026 CPython 3.12musllinux: musl 1.2+ x86-64

sketchlog-1.2.3-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (247.6 kB view details)

Uploaded Jun 30, 2026 CPython 3.12manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

sketchlog-1.2.3-cp312-cp312-macosx_11_0_arm64.whl (221.5 kB view details)

Uploaded Jun 30, 2026 CPython 3.12macOS 11.0+ ARM64

sketchlog-1.2.3-cp312-cp312-macosx_10_13_x86_64.whl (229.4 kB view details)

Uploaded Jun 30, 2026 CPython 3.12macOS 10.13+ x86-64

sketchlog-1.2.3-cp311-cp311-win_amd64.whl (397.3 kB view details)

Uploaded Jun 30, 2026 CPython 3.11Windows x86-64

sketchlog-1.2.3-cp311-cp311-musllinux_1_2_x86_64.whl (1.3 MB view details)

Uploaded Jun 30, 2026 CPython 3.11musllinux: musl 1.2+ x86-64

sketchlog-1.2.3-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (244.7 kB view details)

Uploaded Jun 30, 2026 CPython 3.11manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

sketchlog-1.2.3-cp311-cp311-macosx_11_0_arm64.whl (219.4 kB view details)

Uploaded Jun 30, 2026 CPython 3.11macOS 11.0+ ARM64

sketchlog-1.2.3-cp311-cp311-macosx_10_9_x86_64.whl (226.6 kB view details)

Uploaded Jun 30, 2026 CPython 3.11macOS 10.9+ x86-64

sketchlog-1.2.3-cp310-cp310-win_amd64.whl (396.0 kB view details)

Uploaded Jun 30, 2026 CPython 3.10Windows x86-64

sketchlog-1.2.3-cp310-cp310-musllinux_1_2_x86_64.whl (1.3 MB view details)

Uploaded Jun 30, 2026 CPython 3.10musllinux: musl 1.2+ x86-64

sketchlog-1.2.3-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (243.2 kB view details)

Uploaded Jun 30, 2026 CPython 3.10manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

sketchlog-1.2.3-cp310-cp310-macosx_11_0_arm64.whl (217.9 kB view details)

Uploaded Jun 30, 2026 CPython 3.10macOS 11.0+ ARM64

sketchlog-1.2.3-cp310-cp310-macosx_10_9_x86_64.whl (225.1 kB view details)

Uploaded Jun 30, 2026 CPython 3.10macOS 10.9+ x86-64

File details

Details for the file sketchlog-1.2.3.tar.gz.

File metadata

Download URL: sketchlog-1.2.3.tar.gz
Upload date: Jun 30, 2026
Size: 136.9 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.13

File hashes

Hashes for sketchlog-1.2.3.tar.gz
Algorithm	Hash digest
SHA256	`eef84400df55676cf40c78b03b008b9295236dfb051d058d720a9f3841d62dcb`
MD5	`13aa8c7217a22fa40325ca3cb5911e18`
BLAKE2b-256	`5b4abb9d8a2b3c5c70bb5e669008e859ebbe97c11bcd28ef5d8dea4682b9cf8c`

Algorithm	Hash digest
SHA256	`340bc17f1c867abcf1987f45dd99f263bf22fb69323764c424766381e3163398`
MD5	`9da63227d051218259a1e91417ba9892`
BLAKE2b-256	`186cce465a9c025d0b953f0265ea2be79ee99b5fcb66b906755e8c6b101c863a`

Algorithm	Hash digest
SHA256	`bca6a01e6c149c59fa24dd2c02b4581c1545d4862123662c260e6a4bf96e4879`
MD5	`7fd4c1e71d40036faca0b742462aa192`
BLAKE2b-256	`ced53c0fa7ceb309983bc1cceec6955d9dd43fb31df2b8b7d4dc17f4d86d94d4`

Algorithm	Hash digest
SHA256	`9af44597652c12455cf738bf5cce97c4bac82b96c8f9af602bdac026abdfa2ad`
MD5	`77c3b1a34a34c2936d0ba501147202a9`
BLAKE2b-256	`f1d32d9ebd56d0a1e0f305ee07b8e448bbf0048a8d2658d7591e1b8d71c4f40e`

Algorithm	Hash digest
SHA256	`5c791a9ce955903c2890cf4503ef1dab98a81f5c0444dfee5a708741a1208a2f`
MD5	`01fa0f62896a3f98b605ca4365e78a5b`
BLAKE2b-256	`fea80b77fa4293aa29798d72a008df00d93bf3fcca688f091278ff3876655e40`

Algorithm	Hash digest
SHA256	`e5b7f85a0228e8da37ad0ffa5263e32d3f87026bd8fe2457c04db26f4d414cc9`
MD5	`417bfe25388a0b49f0e006aac1bf7f31`
BLAKE2b-256	`73110eadf9efa9c7e100ea98d0416eb7b2f1dc79c15e3beb1ccb0d6995d320bb`

Algorithm	Hash digest
SHA256	`f79aadaeedfec8c3cf795576fe5ffc7dacee70c28e308e1fd7720526468e7307`
MD5	`b751cab21d9557f1490f1225a81f9732`
BLAKE2b-256	`6941fcbb112f46ab5f6754e07f37b26312329205c599d4e500f078bf30c32a6c`

Algorithm	Hash digest
SHA256	`9a177fc6f591665c769991a7c18fca509ac77d3fc261d2fbe5ffaad15e0dfce3`
MD5	`d3374a557f7df2ebc1147afa4dd48dbb`
BLAKE2b-256	`0bca4b1c4b8818a0b6f8882933e3759a9da35d4ab4592b2d5a6df51e8d9707b1`

Algorithm	Hash digest
SHA256	`07cd672e27efd240f36a026776cfa714d806b0caa5ea495ad65f833424203717`
MD5	`6843d77b1076821276f2fb9b650c72a1`
BLAKE2b-256	`863d20e3fe4603ffe92a7a53817ec1207f4e71c548b43d7e8f1b1e12ccca3777`

Algorithm	Hash digest
SHA256	`1801da9e0260835e2dbf1144c40f71bef9cfd18d30fdd22b79efc76cc6582853`
MD5	`cfb2bcacbcfcb0bf8eff1c1759bac493`
BLAKE2b-256	`03fdb4101f5c4188dc1b9899e28c9f34d0f9266160b8c98b084f0d0d6ec77fbb`

Algorithm	Hash digest
SHA256	`dcd660eb3d22ca7faca6b8ce2736ab1b4f4de121be7c91f25a403da7ae4f84d2`
MD5	`0c34914e074effff3d8673a19f2203b5`
BLAKE2b-256	`af8b936024b69a475184d2f789b78106081c26aa839d9459a50a103ec48105d8`

Algorithm	Hash digest
SHA256	`810478bb4e07bfee3ce0593ff78a5facb6ee4ca978737ace4bd1d280f2afade0`
MD5	`394f426e485418c70d26459ba27a3ab8`
BLAKE2b-256	`564e8e103c10503b1b955cfd3e989d995a43df1541f4a000f728ab82c3722c90`

Algorithm	Hash digest
SHA256	`9c2a718287475411ac23d136d37767c11eac0877ad2486c1fdd1a60381755e33`
MD5	`693111b7f73b2ae7daa201efe8a1ea52`
BLAKE2b-256	`936b71563bae51ed81374211ab09eac5526e3c4099275e11bc7931a55469bdb2`

Algorithm	Hash digest
SHA256	`e27165acdd61b1de3c5abae8a3d9448e6f89ae40b2ab4d1efd65d5bb37ee0e71`
MD5	`6382ff36da9cf48ce4c53bc80c06a614`
BLAKE2b-256	`7685a21c3d1b2f9e0341f594d307c43544462612414f06498c72e3eeb1147494`

Algorithm	Hash digest
SHA256	`2769f000d3dee54d6d74f9af7b7de89b73176f99cddd27e6b49dff682e59fa8d`
MD5	`c40f9647aacbd2d3ce8ff5e7f35e4172`
BLAKE2b-256	`3823844f8479482c52aa535125e4084d485f78015b3fe0f56bf7da4350c91f22`

Algorithm	Hash digest
SHA256	`d1098e56f56719fbb4772a1a693fd3f84d9c436f637d24b161a97d358c83b4c3`
MD5	`d034e5e98b099262b81e2dbd69e8fae6`
BLAKE2b-256	`825f21bb2c46af9f51e9aed0a28cf60943ae24f9d0033753573b83602c25b845`

Algorithm	Hash digest
SHA256	`4842a8f7e864f9f662da469b84f87b7e2fc879edd4b95edb0683fa00422b2ea9`
MD5	`d9f84c240442b985b4296b2b43d7ffb6`
BLAKE2b-256	`bcc51734ee23c7e19865a04a2d82f0f91314d5d9e18209d3f88ac47955a7c040`

Algorithm	Hash digest
SHA256	`3eb472601477f0233d8407102162033485f82ab0bb1d2188b7c2bea8335b5017`
MD5	`9f890f2b36a8c7e401cd532bcc34e293`
BLAKE2b-256	`02f696151da70ce5f1af6f6a0e5abf969c443cede4bc0096d5bceb06d2c5a891`

Algorithm	Hash digest
SHA256	`e8bef37d38b37bbc8f523fa278b35145b2e191ae2f0c279c3c34b92f4d305ff0`
MD5	`4f1de46eb42b1a6188ea2003b33b8646`
BLAKE2b-256	`f90c66d78db9ecac31b9c7aa840edd5ccc0654cf074518b6c588f39895b331f5`

Algorithm	Hash digest
SHA256	`fea90225fe8ba26c4466574b588a38dff4db5b7a4c39fd72acc2460556a72125`
MD5	`769cd4e6a4a915ed0c4003216e6a1f5a`
BLAKE2b-256	`5534ae8a8ab97b83d86ded33f4c68ef80425c0ce3c6a7871e05e60d1b38b0593`

Algorithm	Hash digest
SHA256	`68945eb34dc3c8c27ecdaf4cf98fdefe9f9eaeacb76522bc2ba64c27d128cd2b`
MD5	`5772e49f4b034967a43859d0ab91fd44`
BLAKE2b-256	`85913f1568e51741049e2007b3181090f5ae58ccf5f5bc7d2aa5573a8a8a54d4`

sketchlog 1.2.3

SketchLog

📚 Documentation

Local Documentation Build

Installation

1. The Core Library (Python)

2. The Standalone Server

Quickstart

🐍 Python (Embedded Library)

🟨 TypeScript / Node.js SDK

🐹 Go SDK

Distributed Clustering (Beta)

Community

What this is not

Project links

Key dates

1 maintainer

Credits

GitHub Statistics

License expression

Requires

Provides Extra

Tags

Classifiers

Download files

Source Distribution

Built Distributions

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

Algorithm	Hash digest
SHA256	`5c2123a5e25a058a5a7da7da76012e5e2561a8eecd909cd202aa47502e704680`
MD5	`5cf7bee347a7073385b6b424697e9abe`
BLAKE2b-256	`0704ba61e800d165a90c495d3e7a7dcd69c69730223c26e04907ce273223b30e`

Algorithm	Hash digest
SHA256	`78dec32161f213b1c42544d787f675168ffc1ac11dcf7f4feb3699ceb4f9c712`
MD5	`457e82aa26c3da9ae924d7cddbb55add`
BLAKE2b-256	`37918256e8eae17fe3f7d5bb60f42b78bb72afa50e74fd51a9345c7c0c9f1cea`

Algorithm	Hash digest
SHA256	`cb95fd15809a29b8006d33cb70777bb22c0ec586ccc8441445b02226290f39b3`
MD5	`dd99e5a04c50758cf5d0050174080963`
BLAKE2b-256	`e315a731a85284e0af3f1717e5784f28b936cbbb8ff3f00110ad9c38107f1146`

Algorithm	Hash digest
SHA256	`56c160533be10a7241a947f7c3094c4ce3afde0b5a933da30ccbe92464360d59`
MD5	`0d5ca08c0dc551b3dbeb2e4a15ee2fc9`
BLAKE2b-256	`7d5938149bf1a3f908795e759fc241d61412e1e797d13579a105d4831b1989e1`

Algorithm	Hash digest
SHA256	`6dd3452560edf39088dda7b04f6a66be0123a55d6dbe48eccc52eb5f94c2cc70`
MD5	`c2bd87a9d3d70548c2eb16c7b7ae010f`
BLAKE2b-256	`3ff7d3b3bafa1bc52e7346c29f02f8d87d797a0ce970b3a2df5d9d3ad7388acb`